Abstract:
The purpose of the research. The research is aimed at identifying the possibilities of using various modeling tools in the framework of the functioning of “smart” petrochemical production. To achieve the goal contributed to a number of tasks: to study the possibility of applying IDEF0 methodology in terms of building a “smart” production, to build the decomposition process, the introduction of new equipment in a “smart” the petrochemical industry in BPMN notation, to explore the specific neural network modeling and to develop a neural network model to predict energy consumption of petrochemical plants. Results. It is summarized that modeling is an integral element of the design and management of “smart” petrochemical production, provides process optimization, rationalization of the information and communication environment of the enterprise, energy saving, improving the quality of petrochemical products, production efficiency, reducing the negative impact on the environment; “smart” production should be accompanied by the development of a mechanism for digitalization of production processes, regulation of the algorithm for building cyberspace. We need an integrated approach to production modeling, which involves end-to-end data management at all stages of production, using various modeling methods, and consolidating the results of modeling in a single database, which serves as a relevant empirical basis for making rational management decisions. The study developed a scheme to enter the production assets to the smart production (BPMN notation), a flow of the process, taking into account the difficult situation related to the operation of production equipment; a logical and informational model of the formation of a “smart” petrochemical production (in the IDEF0 notation) is constructed, which takes into account the relationships between sub-processes and the potential effect, allowing for the development of instructions and methodological materials for the modernization of petrochemical production; A predictive model for regulating the level of energy consumption by petrochemical enterprises, depending on the costs of technological innovations and on the volume of polluting emissions generated by industrial enterprises, is proposed, based on training neural networks of different architectures and allowing, in accordance with the activation function of the hidden layer of the neural network, to identify trends in energy consumption.
